On 19.02.18 16:30, Joe Clarke wrote: > Speaking as a contributor... > > On 2/19/18 05:03, Eliot Lear wrote: > > All, > > > Mark Nottingham and I had some extensive conversations about how > > to handle the MUD URL. The issue is was that Mark identified a > > potential violation of a best practice in the way the URL was > > structured. The principles I seek to maintain are these: > > > * The base MUD URL is formed solely by the Thing > > _Formed by_, but relies on the URL pointing to something in the manuf. > world. Meaning, the Thing doesn't just make this up, which is how it > initially read to me. > > > * It may be necessary in the future for the MUD controller to > > provide additional information to the MUD file server. * How that > > happens can be decided in the future, but we should be conservative > > in our approach to leave options opened. > > This sounds reasonable. > > > > Therefore, the result of the discussions is that we should permit > > a normal HTTPS URI as a MUD URL, with all the associated semantics, > > with the exception that queries for now will not be allowed. > > > At the same time, as we build out future work, we will hold > > discussions with the Web/HTTP folk on the best way to proceed. > > There are at least two ways to tackle this in the future: an HTTP > > header, or some limited use of the query component. We needn't > > decide any of that today. > > > Is this okay with you? > > Yes. It is good to be careful as to ensure a strong standard. But... > > Speaking as a co-chair... > > Is this a case of perfect being the enemy of good? I know of some > people working to implement this. I don't think they are all on this > list. Have you run this by them to see if this limits any use cases > out of the box?
I have. I think they're okay with it. They would have been better with it if we had resolved this a year ago but one doesn't get everything one wants. They DO expect us to hold discussions with the apps folks so we can get on to the next thing. Eliot > It wouldn't make sense to put something out that > wouldn't be implemented (i.e., implementors all need that future work > to happen first). > > Joe >
signature.asc
Description: OpenPGP digital signature
_______________________________________________ OPSAWG mailing list [email protected] https://www.ietf.org/mailman/listinfo/opsawg
